Attribute Reduction Of Large-scale Dynamic Covering Information Systems

To deal with uncertainty information,Zakowski generalized Pawlak's rough sets to covering rough sets,which has been widely used in many fields such as knowledge acquisition and feature selection.In practice,there are a large number of dynamic covering information systems,but it is time-consuming to calculate approximations of sets and attribute reduction with the non-incremental approaches.To improve the performance of knowledge discovery,this thesis provides the incremental methods for computing approximations of sets and attribute reduction of dynamic covering information systems as follows:Firstly,we briefly introduce the background and significance of rough sets,and the current research results,the concepts of Pawlak's rough sets and covering rough sets,and provide the motivation of this study.Secondly,based on the characteristic matrix,we study the matrix representation of the upper and lower approximations of sets in the dynamic covering approximation space with the variations of object sets,and provide the incremental method of cal-culating the upper and lower approximations of sets,and illustrate how to calculate the second and sixth upper and lower approximations of sets in the dynamic covering approximation space with several examples.Finally,we propose the related family-based incremental method of calculating the attribute reduction of dynamic covering decision information system with variations of multiple attributes,the incremental algorithm for computing the attribute reduction in dynamic covering decision information system is proposed.Especially,we design the heuristic algorithm for computing attribute reduction in dynamic covering deci-sion information systems,and the experiment results are employed to illustrate the effectiveness of the heuristic algorithm.In summary,this thesis provides an effective incremental method for knowledge discovery in large-scale dynaunic covering information systems,and enriches the theory of covering rough sets.